WHITEPAPER

BETTER ADHERENCE, HIGHER

ENGAGEMENT: FINDING THE

CLINICAL TRIALS SWEET SPOT

SPENCER

HEALTH

SOLUTIONS

Alan Menius, Chief Scientific Officer,

Spencer Health Solutions

Aiden Flynn, Chief Executive Officer,

Exploristics

INTRODUCTION

Clinical trial sponsors and clinical research

organizations must work through an ever-

expanding set of challenges to ensure their

studies of new drugs produce valid results. And

they must do so while strictly managing costs

and timelines. As more trials make use of in-

home settings, digital technologies and real-

time patient data, the pressure increases to

efficiently manage a process that’s dramatically

different from the traditional methods of direct

observation inside the walls of a clinic.

Sponsors are increasingly considering new

technology to help them manage trials in this

fast-developing environment. Technology

platforms, digital applications, electronic

patient diaries, wearables, data sharing,

videoconferencing and various forms of

artificial intelligence are among the many new

methods with the potential to reduce time and

cost associated with trials.

And yet, adoption isn’t where it could be.

According to the Deloitte Center for Health

Solutions, the industry has been slow to update

clinical trials processes with digital capabilities.

This is now changing as organizations see

the potential for patient-centric digital

technologies to transform clinical

development.

The increasing focus on modernizing clinical

trials raises questions about where sponsors

should invest their resources. Is it in tools to

expedite patient recruitment and retention?

Improve data collection and reporting?

Increase patient engagement? Reduce non-

adherence?

Adherence has asignificant impact onpatient retention.

Non-adherence and lack of engagement are

known to play a significant role in whether a

patient finishes a study or drops out. With

respect to patient engagement, increased

engagement between physicians and trial

participants has been found to have a

significantly positive correlation with

adherence. Thus, sponsors are striving to

utilize different strategies and technologies

to enable participants to accurately report

their health status.

Medication adherence is one of many

essential considerations for clinical trial

sponsors and CROs as they move from study

design and patient recruitment to the study

itself and data analysis. Non-adherence in

particular has been shown to increase

variance, lower study power, and reduce the

magnitude of treatment effects.

1

23

2

THE ROLE OFADHERENCE ANDENGAGEMENT INCLINICAL TRIALS

Much is at stake: By reducing non-adherence

by just one percentage point, sponsors can

save an estimated $335,000 in the total cost

of a Phase III trial as a result of needing to

recruit and retain fewer participants to

preserve statistical power.

Causes of non-adherence are diverse but

include lifestyle disruption, side effects, or

they may simply forget to take their

medication. When participants are actively

engaged during the medication process and

take their medications as prescribed during

the trial, it reduces a host of worries for

sponsors.

Simulations are important and relevant to

clinical trials. The FDA promotes modeling

and simulation as an advancement in the

clinical trials process that can help bring

therapies to market more quickly and

effectively. In fact, the agency regularly

advises industry to use modeling and

simulation to predict outcomes from clinical

trials, inform trial design, and support

evidence of effectiveness, among other

reasons.

MEDICATION ADHERENCEDEFINED

The World Health Organization defines

adherence as the degree to which patients’

behavior aligns with the recommendations of a

health care provider. When patients comply

with those recommendations at least 80

percent of the time, they’re considered

adherent.

PATIENT ENGAGEMENTDEFINED

The Patient-Centered Outcomes Research

Institute (PCORI) has established a definition

of patient engagement for clinical research:

“The meaningful involvement of patients,

caregivers, clinicians and other healthcare

stakeholders throughout the entire research

process – from planning the study, to

conducting the study, and disseminating

study results.”

4

5

3

Concerns include the quality of the results,

the cost of the study and the extra patients

needed to be enrolled to replace any

dropouts.

To assess the impact that medication

adherence plays on clinical trials,

Exploristics, a trial design and analysis firm,

partnered with Spencer Health Solutions to

conduct hundreds of simulated

hypertension studies to measure the

impact adherence may play on a

hypertension clinical trial.

The Spencer Health Solutions-Exploristics Simulation:

Modeling a Hypertension Clinical Trial

Clinical trial simulation involves the use of a

model to describe a process or system, executing

the model, and analyzing the outputs. Simulation

is useful when there are multiple, interrelated

factors that impact the outputs.

The Spencer-Exploristics simulation was

designed to estimate the impact of adherence on

the probability of success in a trial of a

hypertensive population with a high risk of stroke.

A stroke study, published in The Lancet, was used

and represented a a randomized trial of 6,105

individuals with previous stroke or transient

ischemic attack. This study was chosen because

hypertension is a chronic condition and this study

allowed for reasonable assumptions to be made

regarding the impact of adherence on stroke.

The underlying model for the simulated studies

comprised the following factors:

4

The simulations enabled an evaluation of the

performance of different study designs under

different compliance scenarios. The scenarios

included studies with sample sizes ranging from

500 to 1,900 patients equally allocated to an

active treatment arm and a control group. Ten

scenarios relating to compliance were defined by

the combination of the proportion of subjects

with non-adherence and the extent of non-

adherence. Results from three of the scenarios are

reflected in the charts on the following pages.

For each compliance scenario and sample size,

500 simulations were run by generating virtual

patient-level data that conforms the assumptions

defined by each scenario. Once the data was

generated, the proportion of patients with stroke

in the active treatment group was compared with

the placebo group using a Chi-squared test and

the comparison was declared significant at the

5% level. The percentage of simulations that

achieved significance then gave the probability of

success for each scenario. The target sample size

for each compliance scenario was determined by

the value that achieved a probability of success

between 80 and 90%.

Risk of stroke in placebo control group

Risk of stroke in treated population

Proportion of subjects with non-adherence in a

study

The extent of non-adherence in the clinical

trials population

Risk of stroke in subjects given their adherence

FINDINGS

After all simulations were completed, two findings stood out:

For example, at an adherence rate of 70 percent, 10.3 percent of trial participants would be

expected to experience a stroke within four years. But when the adherence rate climbs to 90

percent, just 9.1 percent of patients would be expected to experience a stroke. (The placebo

stroke rate after four years was expected to be 14.4 percent.)

First, changes in adherence make a substantial difference in

patient outcomes.

5

*Simulated results. Error bars represent entire range of simulation.

The graph below summarizes the probability of stroke for all 500 simulations with increasing levels of adherence

compared to the published placebo control and treated results from the original study.

Low adherence increases the likelihood of Stroke by 42%

MEDICATIONADHERENCE:

A SIGNIFICANTINFLUENCE

6

Low adherence requires an additional 40% increase in patients needed for a

successful trial.

White numbers represent average number of patients needed for a successful clinical trial at 80% statistical power.

Error bars represent entire range of simulation.

Second, low adherence dramatically increases the number of trialparticipants needed to achieve valid results.

These findings demonstrate that adherence accounts for a larger role in trial efficacy and likely

outcomes than is generally believed. If an increase in adherence of 10 to 20 percentage points

means fewer participants need to be recruited and engaged to completion, the implications on costs

and timelines can be substantial.

For example, an analysis published in Applied Clinical Trials estimated the average cost of enrolling

one patient in a Phase III trial at $26,000.

Thus a typical hypertension study may realize up to an additional $10.3mm in patient

enrollment expenses when comparing the low to high adherence scenario.

The chart below shows that 1,183 subjects would need to be recruited and enrolled at a 70 percent

adherence rate (for 70 percent of participants) . At 90 percent adherence (for 90 percent of

participants) , the number of subjects needed falls to 982, a decrease of 201 patients needed for

recruitment.

SUMMARY

ADHERENCE AND ENGAGEMENT: TAKING CENTER STAGE

The findings from our simulations show that

adherence has a substantial impact on the cost,

efficacy and likelihood of success in clinical

trials. For example, the number of patients

needed to complete a study with a statistical

power of 80% nearly doubles when non-

adherence is 40 percent , compared to full

adherence. These findings have important

implications for sponsors as they search for the

right technologies and methodologies to

modernize their trials.

The transformation of clinical trials is here to

stay. Traditional research and development

processes are being modernized, the

regulatory framework is embracing digital

technologies, and – perhaps most important –

patients are being given a real voice as

stakeholders in clinical trials. These changes

are necessary as pharmaceutical, device, and

life sciences companies face intense pressure

to deliver safe, effective therapies at a greater

value to consumers.

95%medication adherence

81%patient engagement

Patients using a spencer device in the home showed a 95% medication adherence rate of

medications taken during time prescribed. Those same patients provided feedback through the

device 81% of the time when asked a question, averaging 2.5 questions daily. Measures taken 2017-

2019 from pilot with 160 patients (average age 70 and each with at least one chronic condition)

using spencer device in the home.7

CASE STUDY: SPENCER HEALTH SOLUTIONS

Citations:

1 Deloitte Center for Health Solutions: https://www2.deloitte.com/insights/us/en/industry/life-

sciences/digital-research-and-development-clinical-strategy.html

2 Med Care: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2728700/

3 Journal of Clinical Pharmocology: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5066799/

4 Applied Clinical Trials: http://www.appliedclinicaltrialsonline.com/non-adherence-direct-

influence-clinical-trial-duration-and-cost

5 FDA: https://www.fda.gov/science-research/about-science-research-fda/how-simulation-can-

transform-regulatory-pathways

Advancing research and healthcare from the home. Spencer Health Solutions, Inc. leads the digital health

market with the spencer® Smart Hub. Bring new treatments to market faster and at a lower cost with the

award-winning spencer, which combines medication dispensing, telehealth and engagement, so patients,

their healthcare providers and clinical research teams stay connected.

About Spencer Health Solutions

Copyright © 2019 Spencer Health Solutions, Inc. All rights reserved. | www.spencerhealthsolutions.com

| 866-971-8564 | info@spencerhealthsolutions.com

Exploristics provides analytics, statistics, exploratory data analysis, modelling and simulation services. They

have unique expertise in handling large data resources, including Clinical Trials,Biomarker,

Pharmacogenomic, Imaging and Observational data.

About Exploristics

AUTHORS

Alan Menius is the Chief Scientific Officer at Spencer Health Solutions. He is responsible for

developing the scientific strategy for using the spencer SmartHub in clinical research as

well as leveraging spencer data. Prior to this role, Alan spent 25 years at GlaxoSmithKline

where he led advanced analytics teams responsible for leveraging disparate data sources

and advanced analytics to inform the safety and effectiveness of medicines. Alan has over

20 publications in peer-reviewed journals and has given numerous invited presentations

on advanced analytics at national and international meetings.

Aiden Flynn is the founder of Exploristics and has more than 20 years’ experience in

drug discovery and clinical development. After seven years as a lecturer at

University College, London, he spent ten years at GlaxoSmithKline as a director of

statistical support for biomarker studies across research and development. He was

responsible for developing and implementing the pharmacogenetics strategy that

enabled the use of genetics data in clinical studies. This involved the integration of a

wide range of capabilities across R&D such as new analysis tools and methods,

bioinformatics, data standards, processes and training. In recent years, Aiden has

worked closely with regulatory authorities, such as the FDA and EMEA, to develop

tools and guidelines that support the use of biomarkers in clinical studies.

8